The objective of the thesis was to identify Heavy Vehicle (HV) characteristics that have an impact on freeway throughput at various congestion levels with an emphasis on operations at LOS E and F. Furthermore, it was desired to use the available database in order to derive Passenger Car Equivalent (PCE) factors obtained under congested and severely congested conditions in order to compare them to the HCM 2000-recommended PCE factors that were calibrated under βsteady-flowβ conditions.
Since PCE factors are based on the ratio of Heavy Vehicle to Passenger Car (PC) headway ratio per equation 2.1 below, the present research effort focused on factors affecting HV and PC headways at various congestion levels.
PCE= βπ‘
βπ 2.1
Headway relation with speed
A fundamental Headway-Speed quadratic equation (see equation 5.1 below) was fit to the available data and calibrated at the aggregate analysis level, for average headways across all analyzed vehicle classes. A high correlation between the dependent and the independent variables was identified (r = 0.981); the model had a very good fit at the 0.000 significance level.
ππβππππ π»ππππ€ππ¦ = 0.0018 β πππππ2+ β0.14 β πππππ + 4.94 5.1
Where, Vehicle Headway was measured in seconds.
Minimum headways (operation at capacity) were associated with speeds ranging between 40 and 45 mph and densities slightly higher than 45 veh/mi (LOS F), as shown in Figure 5.3.
Recognizing the overarching relationship between headways and speed, the remainder of the data analysis focused on identifying additional factors affecting HV and PC headways and especially factors that may have a differential effect on these two types of headways.
Lagging Passenger Car Headways
Researchers had previously established that PC drivers maintained longer headways when following HV than when following lighter vehicles. Two types of βlight vehiclesβ, PC and light trucks (vehicle classes 2 and 3 respectively-see Figure 3.1 for definitions) and two types of βHVβ (vehicle classes 8 and 9) were used to verify whether this finding applied to the study database. An additional dimension in this investigation was a separate analysis for each of nine 5-mph speed ranges. It was indeed verified that for each analyzed speed range the headways PC drivers maintained from βlight vehiclesβ were statistically indistinguishable; the headways they maintained from class 9 vehicles were statistically significantly longer (Tables 5.6 and 5.7 and Figure 5.5). Although, their headways from class 8 vehicles were longer that those they kept from βlighter vehiclesβ they were not always statistically significantly longer.
A clear pattern of PC drivers maintaining shorter headways from vehicles class 9 as speeds increased from 10-15 mph to 45-50 mph was evident.
Passenger Car-Heavy Vehicle Pair Headways
Contradictory findings about whether the headways PC maintained from HV were longer or shorter than those maintained by HV drivers following PC were addressed using separate headway comparisons with vehicle classes 8 and 9. PC drivers were found to maintain statistically significant shorter headways when following HV than the headways maintained by HV drivers following PC. This finding was valid both for class 8 and class 9 HV and for all analyzed vehicle speed ranges (Table 5.8 and Figures 5.6 and 5.7).
Gross Vehicle Weight-Headway Relation
Headway relations with a general Gross Vehicle Weight (GVW) classification indicated statistically significant larger headways maintained by vehicles with a GVW exceeding 30 kips compared to those with a GVW up to 15 kips (Table 5.9 and Figure 5.8) for all analyzed speed ranges (0 to >50 mph). Heavy Vehicle Percentage and Passenger Car Equivalent Factor Relationship
Five-minute traffic flow summary information was used to identify PC-only periods and the minimum headway during these periods, corresponding to the facility capacity. This headway was compared to PC headways in mixed traffic at four levels of truck presence (>0-3%, 3-6%, 6-9% and >9% trucks in the traffic stream). PC and HV headways (identified separately for vehicles class 8 and 9; also for all vehicles classes 4 and higher) were found to increase with truck presence in the traffic stream (Table 5.10). A similar pattern was found for the
PCE factor, which was shown to be higher than the recommended HCM 2000 value of 1.5 for similar level, basic freeway segments (Table 5.10 and Figures 5.15 and 5.16). This finding did not agree with the currently accepted findings that the PCE factor decreases at higher HV presence levels.
The HV PCE value under congested conditions and with more than 9% HV presence was found to be 1.76 which is higher than the HCM-recommended HV PCE value of 1.5 under steady-flow conditions. Also, Passenger Cars were found to have the effect of more than 1 PC under congested conditions and a high HV presence.
6.2 Recommendations for future research
ο· The lack of adequate HV data operating under severely congested conditions in the analyzed database leads to a recommendation to identify a data collection site experiencing severely congested conditions for many hours each day, used by a significant number of HV so adequate sample sizes will be available to arrive at definitive conclusions and allow the inclusion of the 0-10 mph speed range in the analysis.
ο· It would be desirable that future data collection locations include a vehicle weight collection capability in addition to a vehicle classification capability. It is reasonable to assume, and there are indications in that direction in the analyzed data, that heavy vehicle drivers base their headway choice on vehicle weight. The GVW-headway relation analysis provided indications that vehicle class-based analyses may not adequately account for the overall vehicle weight (GVW) probably due to the payload effect. For
vehicle classes 3-6 maximum payloads can be 50% to 100% of the empty vehicle weight; for semi trucks classes 8 and 9 maximum payload may be up to 200% of their empty weight. Thus a fully loaded class 8 vehicle may be heavier than a half-empty class 9 vehicle; furthermore, empty and loaded vehicles in the same class may maintain substantially different headways, especially at lower speeds.
ο· It would be desirable to analyze vehicle spacing in order to identify vehicle spacing relations with speed, GVW and other parameters. This information will be useful in calibrating separate car-following mathematical models for each vehicle class that would be readily available as inputs for simulation packages.
ο· The present effort analyzed headways measured from front axle-to-front axle. Although this analysis is useful from a freeway capacity analysis point of view, headway and spacing between the rear bumper of a leading vehicle and the front bumper of a trailing vehicle would make more sense from the driver headway/spacing choice point of view.
ο· Future research can analyze the effect of three consecutive heavy vehicles in a lane forming a heavy vehicle βtrainβ in order to test a finding of shorter headways between vehicles forming such βtrains.β
ο· It would be desirable to analyze the effect of heavy vehicle weight/power ratio on headways with field-collected data. Most of currently available information is based on simulated, non-calibrated runs.
BIBLIOGRAPHY
1. American Association of State Highway and Transportation Officials. (2004). A Policy on Geometric Design of Highways and Streets, 2004. American Association of State Highway and Transportation Officials, Washington, DC., 2004.
2. Highway Capacity Manual 2000. TRB, National Research Council, Washington, DC., 2000.
3. Roess, R., and Messer, C. Passenger Car Equivalents for Uninterrupted Flow: Revision of Circular 212 Values. In Transportation Research Record 971. TRB, National Research Council, Washington, DC., 1984, pp. 7-13. 4. Krammes, R., and Crowley, K. Passenger Car Equivalents for Trucks on Level
Freway Segments. In Transportation Research Record 1091. TRB, National Research Council, Washington, DC., 1986, pp. 10-17.
5. West, J., and Thurgood, G. Developing Passenger-Car Equivalents for Left- Turning Trucks at Compressed Diamond Interchanges. In Transportation Research Record 1484. TRB, National Research Council, Washington, DC., 1995, pp. 90-97.
6. Webster, N, and Elefteriadou, L. A Simulation Study of Truck Passenger Car Equivalents (PCE) on Basic Freeway Sections. In Transportation Research, Vol. 33B, 1999, pp. 323-336.
7. Al-Kaisy, A., Hall, F., and Reisman, E. Developing Passenger Car Equivalents for Heavy Vehicles on Freeways During Queue Discharge Flow. In Transportation Research, Vol. 36A, 2002, pp. 725-742.
8. Seguin, E., Crowley, K., and Zweig, W. Urban Freeway Truck Characteristics, Vol. I: Passenger Car Equivalents. Report FHWA/RD-81/156. Institute for Research, State College, Pa., 1981.
9. Sarvi, M. Freeway Operation Characteristics of Heavy Commercial Vehicle Traffic. Presented at 88th Annual Meeting of the Transportation Research Board, Washington, DC., 2009.
10. Tanaboriboon, Y., and Aryal, R. Effect of Vehicle Size on Highway Capacity in Thailand. In Journal of Transportation Engineering, Vol. 116, No. 5. ASCE, 1990, pp. 658-666.
11. Rakha, H., Ingle, A., Hancock, K., and Al-Kaisy, A. Estimating Truck Equivalencies for Freeway Sections. In Transportation Research record 2027. TRB, National Research Council, Washington, DC., 2007, pp. 73-84.